VoxPilot

VoxPilot is an AI-powered voice assistant for YouTube video analysis. It allows users to paste YouTube URLs, get AI-generated summaries, and interact with the content using natural voice commands. The application leverages Google Gemini for intelligent content analysis and ElevenLabs for natural text-to-speech responses.

Features

Video Analysis

• Paste any YouTube URL to analyze the video content
• Automatically extracts transcripts when available for accurate analysis
• Falls back to metadata-based inference when transcripts are unavailable
• Displays confidence badges (Full or Inferred) based on data source
• Generates key takeaways, abstracts, and structured summaries

Voice Control

• Full voice command support for hands-free operation
• Supported voice commands:
  • "Summarize this video" or "Analyze this" to process a pasted URL
  • "Save" or "Save this video" to bookmark the current video
  • "Delete" to remove a video from your library (with voice confirmation)
  • "Read the summary" to hear the summary spoken aloud
  • "How many videos do I have?" to get a count of saved videos
  • "Switch to dark mode" or "Switch to light mode" for theme control
  • "Play" or "Watch" to open the video on YouTube

Video Library

• Save analyzed videos to your personal library
• View and manage saved videos in the sidebar
• One-click loading of previously saved summaries
• Voice commands for library management

Audio Responses

• Natural text-to-speech powered by ElevenLabs
• Voice feedback for confirmations, summaries, and answers
• Confidence-adjusted voice tone for follow-up question responses

Tech Stack

Technology	Purpose
Next.js 14	React framework with App Router and Server Actions
TypeScript	Type-safe development
Tailwind CSS	Utility-first styling
Supabase	Authentication and database for saved videos
Google Gemini	AI-powered video summarization and Q&A
ElevenLabs	Natural text-to-speech for voice responses
Framer Motion	Smooth animations and transitions
Radix UI	Accessible component primitives
Web Speech API	Browser-based voice recognition

Prerequisites

Before running VoxPilot locally, you will need:

1. Node.js (v18 or higher)
2. pnpm package manager
3. Supabase account with a project set up
4. Google AI API key for Gemini access
5. ElevenLabs API key for text-to-speech

Local Development

1. Clone the Repository

git clone https://github.com/your-username/VoxPilot.git
cd VoxPilot

2. Install Dependencies

pnpm install

3. Configure Environment Variables

Copy the example environment file and fill in your API keys:

cp .env.example .env.local

Edit .env.local with your credentials:

# Supabase Configuration
NEXT_PUBLIC_SUPABASE_URL=https://your-project.supabase.co
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key_here

# Google AI API Key (for Gemini)
GOOGLE_API_KEY=your_google_api_key_here

# ElevenLabs API Key (for Text-to-Speech)
ELEVEN_LABS_API_KEY=your_elevenlabs_api_key_here

4. Set Up Supabase Database

Create the following table in your Supabase project:

create table saved_content (
  id uuid default gen_random_uuid() primary key,
  user_id uuid references auth.users(id) on delete cascade,
  url text not null,
  video_id text not null,
  title text not null,
  summary_json jsonb not null,
  thumbnail_url text,
  created_at timestamp with time zone default now()
);

-- Enable Row Level Security
alter table saved_content enable row level security;

-- Create policy for users to manage their own content
create policy "Users can manage own content" on saved_content
  for all using (auth.uid() = user_id);

5. Run the Development Server

pnpm dev

Open http://localhost:3000 in your browser.

Project Structure

VoxPilot/
├── app/
│   ├── actions.ts          # Server actions for AI processing
│   ├── dashboard/          # Main dashboard page
│   ├── login/              # Login page
│   ├── signup/             # Signup page
│   ├── globals.css         # Global styles
│   ├── layout.tsx          # Root layout
│   └── page.tsx            # Landing page
├── components/
│   ├── ui/                 # Reusable UI components
│   └── icons.tsx           # Icon components
├── lib/
│   ├── supabase/           # Supabase client configuration
│   └── utils.ts            # Utility functions
├── types/                  # TypeScript type definitions
└── middleware.ts           # Auth middleware

Usage

1. Sign up or log in to access the dashboard
2. Paste a YouTube URL in the input field
3. Click Analyze or say "Summarize this video"
4. View the AI-generated summary with key takeaways
5. Save videos to your library for later reference
6. Use voice commands for hands-free interaction

Voice Commands Reference

Command	Action
"Summarize this" / "Analyze this"	Analyze the video URL in the input field
"Save" / "Save this"	Save current video to library
"Delete"	Delete current video (requires confirmation)
"Read the summary"	Read summary aloud
"Read the answer"	Read the last Q&A answer aloud
"How many videos"	Count saved videos
"Dark mode" / "Light mode"	Switch theme
"Hello" / "Hi"	Greeting response
Ask any question	Get contextual answers about the video

API Keys

Google Gemini

1. Visit Google AI Studio
2. Create or select a project
3. Generate an API key
4. Add to GOOGLE_API_KEY in .env.local

ElevenLabs

1. Visit ElevenLabs
2. Sign up and go to Profile Settings
3. Copy your API key
4. Add to ELEVEN_LABS_API_KEY in .env.local

Supabase

1. Visit Supabase Dashboard
2. Create a new project
3. Go to Project Settings > API
4. Copy the URL and anon key
5. Add to .env.local

License

This project is licensed under the MIT License.